CN116431895A - Personalized recommendation method and system for safety production knowledge - Google Patents
Personalized recommendation method and system for safety production knowledge Download PDFInfo
- Publication number
- CN116431895A CN116431895A CN202310131698.XA CN202310131698A CN116431895A CN 116431895 A CN116431895 A CN 116431895A CN 202310131698 A CN202310131698 A CN 202310131698A CN 116431895 A CN116431895 A CN 116431895A
- Authority
- CN
- China
- Prior art keywords
- user
- content
- production knowledge
- model
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 226
- 238000000034 method Methods 0.000 title claims abstract description 94
- 238000002372 labelling Methods 0.000 claims abstract description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 54
- 230000006399 behavior Effects 0.000 claims description 46
- 238000004458 analytical method Methods 0.000 claims description 39
- 238000001914 filtration Methods 0.000 claims description 24
- 230000000694 effects Effects 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 description 20
- 239000013598 vector Substances 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 238000010248 power generation Methods 0.000 description 6
- 238000010606 normalization Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000005065 mining Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000000388 lateral force microscopy Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 238000000513 principal component analysis Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a personalized recommendation method and a personalized recommendation system for safe production knowledge, wherein the method comprises the following steps: determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information; labeling the user interest model and determining a user label model; and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model. The method and the device can accurately and individually recommend interested safe production knowledge content to the user, and avoid the user from browsing useless information.
Description
Technical Field
The invention relates to the technical field of intelligent recommendation, in particular to a personalized recommendation method and system for safety production knowledge.
Background
With the rapid development and popularization of the internet, a personalized recommendation system (Personal Recommender System) has become an important research field as a kind of general information retrieval. The personalized recommendation system can extract a certain rule from the historical data accumulated by the user, automatically filter out some unnecessary information for the user, and actively push the most suitable information according to the preference of the user.
The personalized recommendation algorithm is the most core technology in the recommendation system, so that the performance of the recommendation system is determined to a great extent, whether information really interested by a user can be recommended or not is determined, the recommendation system is required to accurately recommend the information facing the continuously-improved demands of the user, and the recommendation system also analyzes and recommends the latest result in real time according to the behavior of the user, so that accurate intelligent recommendation for the user is realized.
However, mass information is generated on the internet every day at present, and by means of the existing recommendation algorithm, it is more difficult to quickly and accurately find the content interested by the user from the mass information, and the recommendation accuracy is poor. If the user's goal is clear, the search function may be used to actively retrieve the content of determined interest. However, many times, the user does not have an explicit target, and at this time, it is difficult to efficiently match the content of interest to the user, and personalized services cannot be continuously and accurately pushed to the user, so that the user needs to browse a large amount of redundant useless information, and browsing time is wasted.
Therefore, how to provide a personalized recommendation method and a personalized recommendation system for safe production knowledge, which accurately and individually recommend interested safe production knowledge content to users, avoids users from browsing useless information, and becomes a problem to be solved urgently.
Disclosure of Invention
Aiming at the defects in the prior art, the embodiment of the invention provides a personalized recommendation method and a personalized recommendation system for safety production knowledge.
The invention provides a personalized recommendation method for safe production knowledge, which comprises the following steps:
determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information;
labeling the user interest model and determining a user label model;
and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
According to the personalized recommendation method for the safety production knowledge, before the step of determining the recommended content of the target safety production knowledge in the labeled safety production knowledge base based on the user label model, the personalized recommendation method further comprises the following steps:
determining a content tag based on the content information subject and the information keyword in the secure production knowledge base;
and mapping and binding the content information in the security production knowledge base with the content label based on the content label, and determining the labeled security production knowledge base.
According to the personalized recommendation method for the safety production knowledge, which is provided by the invention, the user interest model is determined based on the user data information, and the method specifically comprises the following steps:
Analyzing the user data information according to a preset analysis direction to determine the user attribute; the preset analysis direction comprises the following steps: RFM analysis, user behavior analysis, user preference analysis, and user basic attribute analysis; the user attributes include: user natural attributes, user viscosity attributes, user preference attributes, user collection attributes, user search attributes, and activity policy attributes;
based on the user attributes, a user interest model is determined.
According to the personalized recommendation method for the safety production knowledge, which is provided by the invention, the user interest model is labeled, and the user label model is determined, which comprises the following steps:
labeling the user interest model according to a preset labeling rule, and determining a user label model; the user tag model comprises content information tags of interest to the user and association frequency of the user and each tag.
According to the personalized recommendation method for the safety production knowledge, provided by the invention, the recommendation content of the target safety production knowledge is determined in the labeled safety production knowledge base based on the user label model, and the personalized recommendation method further comprises the following steps:
determining a content list of interest of a user in a tagged security production knowledge base based on a user tag model;
And determining the recommended content of the target safe production knowledge according to the preset content recommendation rule based on the content list of interest of the user.
According to the personalized recommendation method for the safety production knowledge, provided by the invention, the recommendation content of the target safety production knowledge is determined in the labeled safety production knowledge base based on the user label model, and the personalized recommendation method specifically comprises the following steps:
based on the attribute of the content information tag, determining the content of interest of the user in a tagged safe production knowledge base according to a mixed recommendation algorithm; the mixed recommendation algorithm comprises a collaborative filtering recommendation algorithm and a content-based recommendation algorithm;
and ordering the interesting contents of the user according to the association frequency of the user and each label, and determining an interesting content list of the user.
According to the personalized recommendation method for the safe production knowledge, provided by the invention, based on the content list of interest of the user, the recommended content of the target safe production knowledge is determined according to the preset content recommendation rule, and the personalized recommendation method specifically comprises the following steps:
according to the user interested content list and the preset content recommendation rule, calculating the weight of the user interested content;
determining a recommendation candidate list according to the weight of the interesting content;
and determining the recommended content of the target safety production knowledge according to the recommended candidate list.
The invention also provides a personalized recommendation system for the safety production knowledge, which comprises the following steps: an interest model determining unit, a tag model determining unit, and a recommended content determining unit;
an interest model determining unit for determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information;
the label model determining unit is used for labeling the user interest model and determining a user label model;
and the recommended content determining unit is used for determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the steps of any of the above personalized recommendation methods for the safety production knowledge when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of any of the secure production knowledge personalized recommendation methods described above.
According to the personalized recommendation method and system for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, the interested safe production knowledge content is accurately recommended to the user in a personalized way, useless information browsing by the user is avoided, user experience is effectively improved, and user viscosity is enhanced.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a personalized recommendation method for safety production knowledge provided by the invention;
FIG. 2 is a schematic flow chart of a personalized recommendation method for safety production knowledge provided by the invention;
FIG. 3 is a schematic flow chart of a recommendation method provided by the invention;
FIG. 4 is a schematic diagram of the collaborative filtering algorithm provided by the present invention;
FIG. 5 is a schematic diagram of an intelligent recommendation system architecture provided by the present invention;
FIG. 6 is a schematic diagram of a personalized recommendation system for safety production knowledge provided by the invention;
fig. 7 is a schematic diagram of an entity structure of an electronic device according to the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Fig. 1 is a flowchart of a personalized recommendation method for safety production knowledge, provided by the invention, as shown in fig. 1, an embodiment of the invention provides a personalized recommendation method for safety production knowledge, which includes:
step S1, determining a user interest model based on user data information; wherein the user data information includes: user history behavior data and user base information;
step S2, labeling the user interest model, and determining a user label model;
and step S3, determining the recommended content of the target safety production knowledge in the labeled safety production knowledge base based on the user label model.
In particular, in the face of increasing security production knowledge, security production knowledge that users search for through a keyword has grown from tens to tens of thousands, and how to perform information mining on mixed big data through algorithms is a problem that must be faced. The invention provides a personalized recommendation method for safe production knowledge, which is characterized in that user portrait is carried out on a user through user data information, and the personalized recommendation of interesting contents of the user is realized by combining the user portrait with active pushing.
The safety production industry includes a number of different areas of electricity, construction and food production. Further, electric power can be further subdivided into different small fields such as thermal power generation, wind power generation, hydroelectric power generation, power transmission and distribution, and the like. It can be understood that the invention can be adaptively applied to content recommendation (such as agricultural knowledge recommendation, library book recommendation and the like) in other fields besides safe production knowledge recommendation, and the specific application method can be adjusted according to actual requirements, so that the invention is not limited.
When user data information is collected, the front-end embedded point mode is generally adopted to report clicking, searching, collecting and other behaviors of a user, historical behavior data of the user are obtained, and basic information (such as gender, age, region and the like) of the user is obtained. The page data is collected, for example by javascript, and sent to a user data collection server, which stores the data onto an HDFS (Hadoop Distributed File System, distributed file system).
It should be understood that the specific method for collecting user data information is only one specific example of the present invention, and other data collecting methods may be used in practical application of the present invention, which is not limited thereto.
After the user data information is obtained, the data are cleaned and data are mined, and in step S1, characteristics such as user preference, personal information and habit are mined based on the user data information, and a user interest model is built through a scoring matrix model.
For example: based on the user-interest scoring matrix representation, the user interest model is represented by an m x n matrix, where m is the number of system users and n is the user interest. User interest model R ij The interest score of user i for interest j is represented.
Fig. 2 is a flow chart of a personalized recommendation method for safety production knowledge, as shown in fig. 2, in the step of determining a user interest model (part a in the figure), user data information is obtained from a database or a cache, firstly, behavior features of a user are extracted according to user historical behavior data (a user behavior table), and feature vectors capable of representing the user interest are determined by combining with basic information of the user, and the user interest model is determined according to the feature vectors.
After determining the user interest model, in step S2, the user interest model is labeled according to the labeling rule, and a user label model is determined. It can be understood that the specific type of the tag and the tag rule adopted for labeling the user interest model can be set according to actual requirements, and the invention is not limited to this.
The labels corresponding to the users can form the virtual models of the users, and further, the user groups can be subdivided through the labels, and different contents are recommended to the users of different virtual models when the users are individually pushed.
After determining the user tag model, in step S3, based on the user tag model, searching the tagged security production knowledge base for content matching with the user tag model as the recommended content of the target security production knowledge.
It can be understood that each content file in the tagged security production knowledge base should include at least one corresponding tag, and the tag determination method in the tagged security production knowledge base may be determined according to actual requirements, which is not limited in the present invention.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, the interested safe production knowledge content is accurately recommended to the user in a personalized way, useless information browsing by the user is avoided, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for the safety production knowledge provided by the invention, before the step of determining the recommendation content of the target safety production knowledge in the labeled safety production knowledge base based on the user label model, the personalized recommendation method further comprises the following steps:
determining a content tag based on the content information subject and the information keyword in the secure production knowledge base;
And mapping and binding the content information in the security production knowledge base with the content label based on the content label, and determining the labeled security production knowledge base.
Specifically, before determining the recommended content of the target safety production knowledge in the labeled safety production knowledge base by using the user label model, it is necessary to determine to construct the labeled safety production knowledge base.
Based on the information subject and information keywords of the content in the safe production knowledge base, all the content in the safe production knowledge base is marked with corresponding labels (such as industry labels including electric power, water conservancy, building, petroleum, mining and the like, keyword labels including fire prevention, electric shock prevention, high-altitude operation, safety knowledge and the like, and position labels including country, province, city and the like).
It can be understood that the specific implementation method for determining the content corresponding label in the secure production knowledge base can be set according to actual requirements, which is not limited by the present invention.
Furthermore, besides the method of labeling some characteristics of the user by labeling through data mining analysis of clicking, browsing, searching and the like of the user, the user can actively label the interested information content and label the interested information content by focusing attention, collection and the like of the user, so that the matching degree of the label and the interested content of the user is realized.
And mapping and binding the content information in the security production knowledge base with the content labels based on the content labels corresponding to each content, and determining the labeled security production knowledge base.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests and tagging of a safe production knowledge base, quick matching of user interest tags and content tags can be achieved, interested safe production knowledge content is recommended to users in an individualized mode accurately, users are prevented from browsing useless information, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for safety production knowledge provided by the invention, the user interest model is determined based on the user data information, and the method specifically comprises the following steps:
analyzing the user data information according to a preset analysis direction to determine the user attribute; the preset analysis direction comprises the following steps: RFM analysis, user behavior analysis, user preference analysis, and user basic attribute analysis; the user attributes include: user natural attributes, user viscosity attributes, user preference attributes, user collection attributes, user search attributes, and activity policy attributes;
Based on the user attributes, a user interest model is determined.
Specifically, when user interest modeling is performed, according to a preset analysis direction, user data information is analyzed from four analysis directions of RFM (accuracy, frequency, monnetary) analysis, user behavior analysis, user preference analysis and user basic attribute analysis.
In RFM analysis, R represents that the smaller the last data time interval (recovery) R, the higher the user data value; f represents that the larger the Frequency (Frequency) F of the generated data is, the higher the user data value is; m represents the amount of generated data (Montary) M, the larger the user data value is, the higher. The user behavior analysis mainly analyzes the behaviors of the source place, the residence time, the jump rate, the interviewee, the new interviewee, the interview times, the interview interval days and the like of the user. The user preferences mainly analyze relatively stable features that can be represented in the personality of the user, such as which types of websites like to browse, which industries like to pay attention to. User basic attribute analysis mainly comprises analysis of basic attributes such as user age, sex, engaged industry and the like.
Six attributes are designed from the four analysis directions to be classified into a first class, namely a user natural attribute, a user viscosity attribute, a user preference attribute, a user collection attribute, a user search attribute and an activity policy attribute.
It can be understood that, in the practical application of the present invention, the content actually included in the above six attributes can be adaptively adjusted according to the requirements, for example, natural attributes: age, sex, and engaging in industry; viscosity properties: user viscosity refers to the degree of dependence formed by combining trust, benign experience and the like of a user on a website; preference attributes: such as what types of websites the user likes to, the content of the industry, etc. Collection attributes: the collection of the content of the user on the website; search attributes: content searched on the website by the user and searching habit; activity policy attributes: the way the user obtains the security knowledge, such as self-learning, on-line training of organization, etc.
Based on the user attributes, a user interest model may be determined.
It can be understood that when the user interest model is determined, secondary classification can be set under attribute classification, further integration is performed on the user interest features, the most fit requirement mode is selected to perform mixed arrangement, and the corresponding relation between the attributes is integrated.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests and tagging of a safe production knowledge base, quick matching of user interest tags and content tags can be achieved, interested safe production knowledge content is recommended to users in an individualized mode accurately, users are prevented from browsing useless information, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for safety production knowledge provided by the invention, the user interest model is labeled, and the user label model is determined, which specifically comprises the following steps:
labeling the user interest model according to a preset labeling rule, and determining a user label model; the user tag model comprises content information tags of interest to the user and association frequency of the user and each tag.
Specifically, as shown in fig. 2, after determining the user interest model, the user interest model is labeled according to a preset label rule, the feature vector is converted into a corresponding label, and the user label model is determined. The user tag model comprises content information tags of interest to the user and association frequency of the user and each tag. By designing the content information labels and the label frequency, when the recommended content can be determined in the safe production knowledge base, the target is more definite, the pertinence is stronger, and the personalized recommendation can obtain better effects.
For example, the labels of the related content information interested by the user and the association frequency of the user and each label are mined and analyzed by analyzing the frequency of clicking, collecting or browsing labels of the content in the user interest model, browsing time, keyword frequency of the content searched by the user and the like. Specific preset tag rules can be set according to actual requirements, and the invention is not limited to this.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, the user label model is determined through labeling of the model, and the content information labels interested by the user and the association frequency of the labels are determined. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, the problem of overload of safety production knowledge data is solved, passive acquisition is changed into active pushing, interested safety production knowledge content is recommended to a user in an individualized mode accurately, useless information browsing by the user is avoided, browsing time of the user is saved, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for safety production knowledge provided by the present invention, based on the user tag model, in the tagged safety production knowledge base, the recommendation content of the target safety production knowledge is determined, and the method further includes:
determining a content list of interest of a user in a tagged security production knowledge base based on a user tag model;
and determining the recommended content of the target safe production knowledge according to the preset content recommendation rule based on the content list of interest of the user.
Specifically, after the user tag model is determined, the relevant content of interest of the user is screened through the tag in the tagged security production knowledge base through calling the interface, and a content list of interest of the user is determined.
In the tagged security production knowledge base, the specific manner of determining the content of interest of the user may be to perform matching screening according to all tags of the user, or to perform screening only according to tags with a frequency exceeding a certain number, and the like, and may be adjusted according to actual requirements, which is not limited in the present invention.
After the user interested content list is determined, determining target safe production knowledge recommended content based on the user interested content list according to preset content recommendation rules, and pushing the target safe production knowledge recommended content to the user.
It may be appreciated that the preset content recommendation rule (shown in part C of fig. 2) determines a rule of a final push result in the content list of interest of the user, for example, filters recommendation content including forbidden tags (forbidden keywords), calculates similarity between the rest content tags and the user tags, selects contents with the first few names of similarity as the final recommendation result, or ranks according to the total tag frequency, selects contents with the first few names of total frequency as the final recommendation result, and the specific rule may be set according to actual requirements, which is not limited in the present invention.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, related content of the user interests is screened through tags, the problem of overload of safety production knowledge data is solved, passive acquisition is changed into active pushing, the safety production knowledge content of the interests is accurately and individually recommended to the user, useless information browsing by the user is avoided, browsing time of the user is saved, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for safety production knowledge provided by the invention, based on the user tag model, the recommendation content of the target safety production knowledge is determined in the tagged safety production knowledge base, and the method specifically comprises the following steps:
based on the attribute of the content information tag, determining the content of interest of the user in a tagged safe production knowledge base according to a mixed recommendation algorithm; the mixed recommendation algorithm comprises a collaborative filtering recommendation algorithm and a content-based recommendation algorithm;
And ordering the interesting contents of the user according to the association frequency of the user and each label, and determining an interesting content list of the user.
Specifically, based on the attribute of the content information tag, determining the content of interest of the user in a tagged security production knowledge base according to a mixed recommendation algorithm; the mixed recommendation algorithm comprises a collaborative filtering recommendation algorithm and a content-based recommendation algorithm (shown in part B in fig. 2), and the content of interest of the user is ranked according to the association frequency of the user and each label, so as to determine a content list (initial recommendation result) of interest of the user.
Fig. 3 is a schematic flow chart of a recommendation method provided in the present invention, as shown in fig. 3, when determining Content of interest to a user according to an attribute of a Content information tag, a plurality of recommendation engines (recommendation algorithms, such as collaborative filtering recommendation algorithm and Content-based recommendation algorithm (Content-Based Recommendations, CB) and the like) may be adopted to determine the Content of interest to the user together.
Fig. 4 is a schematic diagram of the principle of the collaborative filtering algorithm provided by the present invention, and as shown in fig. 4, the collaborative filtering recommendation algorithm includes the following steps:
1.1 relational matrix and matrix calculation
1.1.1 users and users (U-U matrix)
The use principle is as follows: the Pearson correlation coefficient is mainly used for measuring the correlation between two variables i and j, and the value range is +1 (strong positive correlation) to-1 (strong negative correlation)
Algorithm input: user behavior log.
Algorithm output: based on the collaborative user similarity matrix.
A. And acquiring the relation data between the user and the safety production knowledge content from the user behavior log, namely grading data of the user on the safety production knowledge content.
B. For n users, calculating the similarity between the user 1 and other n-1 users in turn; and then calculating the similarity between the user 2 and other n-2 users.
For any two of users i and j:
a) Searching a safety production knowledge content set which is jointly evaluated by two users;
b) Calculating average evaluation sums of the user i and the user j respectively;
c) And calculating the similarity between the users to obtain the similarity between the users i and j.
C. And storing the calculated similarity result in a database.
1.1.2 safe production knowledge content and safe production knowledge content (V-V matrix)
The use principle is as follows: abstracting the evaluation value of the safety production knowledge content into column vector sum in n-dimensional user space, and using modified cosine similarity
Algorithm input: user behavior log.
Algorithm output: knowledge content similarity matrix is produced based on synergic safety.
A. And acquiring the relation data between the user and the safety production knowledge content from the user behavior log, namely grading data of the user on the safety production knowledge content.
B. For n pieces of safety production knowledge content, calculating similarity between the safety production knowledge content 1 and other n-1 pieces of safety production knowledge content in sequence; and then calculating the similarity between the safety production knowledge content 2 and other n-2 safety production knowledge contents.
C. And storing the calculated similarity result in a database.
1.1.3 Principal Component Analysis (PCA)
In the recommendation system, the dimension reduction can be performed on the security production knowledge content with more attributes by PCA processing, and the m multiplied by n security production knowledge content matrix is converted into a new m multiplied by k matrix.
The object is: the PCA aims to re-map the resulting data space using another set of bases that reveal relationships of the original data as much as possible.
Essence: in effect, the K-L transform, also known as: optimal orthogonal transform
With knowledge of the K-L transform (optimal orthogonal transform) padded, a detailed calculation process for PCA is provided next.
1. Data normalization
Data normalization is performed to eliminate the difference caused by dimension (i.e., unit). Common data normalization methods are 0-1 normalization, maximum and minimum normalization.
2. Calculating covariance matrix
Assuming that the normalized data matrix is X, the covariance matrix cov (X, X) =1nxtx.
3. And obtaining eigenvalues and eigenvectors of the covariance matrix.
There is a corresponding function of eigenvalues and eigenvectors within either matlab or python.
4. Sorting the feature values from large to small, and obtaining corresponding feature vectors according to the principal component contribution rate and the principal component number
1.2 collaborative filtering algorithm based on memory
1.2.1 user-based collaborative filtering algorithms
The collaborative filtering algorithm based on the user mainly comprises the following two steps:
A. historical information of the user and the safe production knowledge content is collected, similarity between the user u and other users is calculated, and a user set N (u) similar to the interest of the target user Ui is found.
B. The safe production knowledge content that the user likes in the collection and that the target user has not heard is found and recommended to the target user.
Applicability(s)
Because the user similarity matrix needs to be calculated, the collaborative filtering algorithm based on the user is suitable for occasions with fewer users; because of strong timeliness, the method is suitable for the field with less obvious personalized interests of users.
Online real-time recommendation
The online service algorithm needs to cache the relevant table in the memory, and then can predict in real time online. In generating the recommendation list to the user, the interest weight of the user for all the safe production knowledge content needs to be calculated and then ranked, and the recommendation list is not suitable for a system with a very large number of safe production knowledge content, if the recommendation list is needed, a relatively fast algorithm is needed to calculate a relatively small candidate list for the user, and then the user is re-ranked by using the LFM (Latent Factor Model, latent semantic model).
On the other hand, LFMs are too slow to calculate online in real time when generating a user recommendation list, and require that the recommendation results of all users be stored in a database offline in advance. Thus, LFM cannot make online real-time recommendations, i.e. the user does not change his recommendation list after having a new behavior.
It will be appreciated that collaborative filtering recommendations fall into 3 categories: collaborative filtering recommendation based on user, collaborative filtering recommendation based on article, and collaborative filtering recommendation based on model. The collaborative filtering algorithm used in the above embodiment of the present invention is a latent semantic model based on LFM matrix decomposition, where LFM appears as an implementation method of the collaborative filtering algorithm. In addition, in the practical application of the present invention, the type of algorithm can be adaptively changed according to the actual requirement, which is not limited by the present invention.
The content-based recommendation algorithm (CB) comprises the steps of:
2.1 basic CB recommendation algorithm
The basic CB recommendation algorithm utilizes the similarity of the basic information of the safety production knowledge content and the user preference content to conduct safety production knowledge content recommendation. And generating preference content of the user by analyzing the safety production knowledge content which the user has browsed, and recommending other safety production knowledge content which has high similarity with the safety production knowledge content which the user is interested in.
Algorithm flow
Algorithm input: knowledge content information and user behavior logs are safely produced.
Algorithm output: and (5) initially recommending results.
A. Secure production knowledge content representation: each secure production knowledge content is represented using a feature vector;
B. the safe production knowledge content set M browsed, collected, evaluated and shared by the user is obtained from the user behavior log, and the content preference of the user can be learned according to the characteristic data of the safe production knowledge content in the safe production knowledge content set M;
C. and storing Top-K safe production knowledge contents into the initial recommendation result.
Applicable scene
The method is suitable for building the basic CB architecture, is particularly effective in that new online safe production knowledge content can be immediately recommended, and the recommended opportunity is the same as that of old safe production knowledge content.
2.2 Linear Classification-based CB recommendation algorithm
When considering content-based secure production knowledge content recommendation as a classification problem, a variety of machine learning methods may be employed. From a more abstract perspective, most learning methods strive to find a linear classification model coefficient that can accurately distinguish between user likes and dislikes of safe production knowledge content.
The safe production knowledge content data is represented by an n-dimensional feature vector, and the linear classifier tries to find a plane capable of correctly classifying the safe production knowledge content in a given safe production knowledge content feature space, wherein one class of points is on one side (like) of the plane as much as possible, and the other class is on the other side (dislike) of the plane.
In order to make up for the advantages and disadvantages of the existing recommendation method, the invention adopts a mixed recommendation mode during recommendation, combines content recommendation and collaborative filtering recommendation, respectively uses a content-based method and a collaborative filtering recommendation method to generate a recommendation prediction result, and then combines the results. One of the most important principles of combined recommendation is to avoid or remedy the weaknesses of the respective recommendation technique after combining.
For example: the collaborative filtering algorithm can find the content A which is interested by other users Li IV and is interested by the user Zhang Sanqi, and then find other content B which is similar to the content A by combining the content recommendation algorithm, and recommend the content B to the user Zhang Sanqi, so that the defect of single recommended content is overcome by combining the content recommendation and the collaborative recommendation filtering, and the recommended content is richer and accords with the user interest.
It should be understood that the above-mentioned combination recommendation method is merely used as a specific example to illustrate the present invention, and the specific combination method can be adjusted according to the actual requirement in the actual application, which is not limited by the present invention.
Further, as shown in fig. 3, in addition to the above-mentioned recommendation of the content of interest to the user using two recommendation engines (collaborative filtering recommendation algorithm and content-based recommendation algorithm), other recommendation engines may be used to recommend other content (such as local information, school news and friends related to the user), and the specific recommendation mode may be set according to the actual requirement, which is not limited in the present invention.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of the user interests, related content of the user interests is screened through tags in a hybrid recommendation algorithm mode, the problem of overload of safety production knowledge data is solved, quick acquisition of the content of the user interests is achieved, accuracy of recommending the content of the interests is effectively improved, passive acquisition is active pushing, the content of the safety production knowledge of the interests is recommended to the user in a personalized mode, useless information browsing of the user is avoided, browsing time of the user is shortened, user experience is effectively improved, and user viscosity is enhanced.
Optionally, according to the personalized recommendation method for safety production knowledge provided by the invention, based on the content list of interest of the user, the recommended content of the target safety production knowledge is determined according to the preset content recommendation rule, and the method specifically comprises the following steps:
according to the user interested content list and the preset content recommendation rule, calculating the weight of the user interested content;
determining a recommendation candidate list according to the weight of the interesting content;
and determining the recommended content of the target safety production knowledge according to the recommended candidate list.
Specifically, after the user interested content list is determined according to the user tag model, the weight of the user interested content is calculated according to the user interested content list and the preset content recommendation rule. It can be understood that the specific calculation formula of the weight can be set according to actual requirements, which is not limited by the present invention.
After the weight of the interesting content is calculated, the interesting content of the user is orderly ordered in sequence, and a recommendation candidate list is determined.
And determining the recommended content of the target safety production knowledge according to the recommended candidate list. It will be appreciated that when the final recommended content is selected from the recommendation candidate list, the number of selected targets may be determined according to the actual requirement, which is not limited by the present invention.
According to the personalized recommendation method for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, relevant content of the user interests is screened through tags, final recommended content which is most fit with the user interests is selected through weight calculation, the problem of overload of safety production knowledge data is solved, the safety production knowledge content which is interesting is automatically recommended to the user in an individualized mode through passive acquisition to active pushing, useless information browsing of the user is avoided, browsing time of the user is saved, user experience is effectively improved, and user viscosity is enhanced.
Fig. 5 is a schematic diagram of an intelligent recommendation system architecture provided in the present invention, and as shown in fig. 5, the present invention is described by taking an example of an intelligent recommendation system as a specific application example of the present invention.
The invention provides a personalized recommendation method of safety production knowledge, which is a tag recommendation method based on UGC (user generated content, content generated by a user), and mainly uses the behavior of tagging the user to recommend related information content for the user, and when the user tags the information content, the user also provides a tag suitable for the content.
The user describes the opinion of the content with tags, which are important data sources reflecting the interests of the user, and the tags are provided to the user in three general ways:
1. recommending the hottest label in a system to a user;
2. recommending labels with the hottest keyword entries in the information content to the user;
3. frequently used tags are recommended to the user.
When each label is generated, firstly, user characteristic description corresponding to the label is required to be defined on the service, then, a user detail attribute group corresponding to the user characteristic label is determined, and finally, a label rule based on the user characteristic attribute group is formulated.
In the embodiment of the invention, the architecture of the intelligent recommendation system can be divided into:
front stage display: the method is that intelligent recommendation, guessing you like, hot spot, watching again, exclusive customization and finding more information and content are seen on the webpage.
Background log system: log data such as user data, user behavior data, content data, etc. is collected, stored, cleaned and analyzed to create a portrayal hierarchy comprising user portraits and content portraits. The user portrait system is generated based on the user tag model and displays the user portrait through the chart. For example: the labels of related content information interested by the user and the frequencies of the user and the labels are mined and analyzed by analyzing the frequencies of clicking, collecting or browsing the labels of the content, browsing time, keyword frequency of the content searched by the user and the like.
Recommendation algorithm engine: various algorithm models, model training configurations and recommended effect evaluation systems.
The intelligent recommendation system comprises three systems: the system comprises a data acquisition system, a data analysis system and a recommendation engine. The data acquisition system is responsible for data acquisition by various standards and private interfaces; the data analysis is responsible for mass data storage and data mining, and the mining results are synchronized to the recommendation subsystem; the recommendation engine provides recommendation data to the external business system.
From a framework perspective, the recommendation system can be basically divided into a data layer, a trigger layer, a fusion filter layer and a ranking layer.
The data layer comprises data generation and data storage, and is mainly formed by cleaning an original log by using various data processing tools, processing the original log into formatted data, and landing the formatted data in different types of storage systems for downstream algorithms and models.
The candidate set trigger layer mainly generates recommended candidate sets from the historical behaviors, the real-time behaviors and the like of the user by utilizing various trigger strategies. The candidate set fusion and the filter layer have two functions, namely, fusion is carried out on different candidate sets, and the coverage and the accuracy of a recommendation strategy are improved; in addition, certain filtering responsibilities are also required, and some manual rules are determined from the aspects of products and operation, and unconditional items are filtered. The sorting layer mainly uses a machine learning model to reorder the candidate set screened by the triggering layer.
Meanwhile, for both the candidate set triggering and reordering layers, the iteration is two layers that need frequent modification for the effect. In order to support efficient iteration, the two layers of candidate set triggering and reordering are decoupled, and the results of the two layers are orthogonal, so that comparison tests can be performed respectively without mutual influence. Meanwhile, in each layer, the flow is divided into a plurality of parts according to the user, and a plurality of strategies are supported to be compared on line at the same time.
It should be understood that the architecture of the intelligent recommendation system is only one specific example of the present invention, and is not meant to limit the present invention.
Fig. 6 is a schematic structural diagram of a personalized recommendation system for safety production knowledge, as shown in fig. 6, and the invention further provides a personalized recommendation system for safety production knowledge, which includes: an interest model determination unit 610, a tag model determination unit 620, and a recommended content determination unit 630;
an interest model determining unit 610 for determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information;
a tag model determining unit 620, configured to tag the user interest model, and determine a user tag model;
The recommended content determining unit 630 is configured to determine, in the tagged security production knowledge base, a recommended content of the target security production knowledge based on the user tag model.
In the face of increasing safety production knowledge, the safety production knowledge searched by users through a keyword has grown from tens to tens of thousands, and how to mine information on mixed big data through algorithms is a problem which needs to be faced. The invention provides a personalized recommendation method for safe production knowledge, which is characterized in that user portrait is carried out on a user through user data information, and the personalized recommendation of interesting contents of the user is realized by combining the user portrait with active pushing.
The safety production industry includes a number of different areas of electricity, construction and food production. Further, electric power can be further subdivided into different small fields such as thermal power generation, wind power generation, hydroelectric power generation, power transmission and distribution, and the like. It can be understood that the invention can be adaptively applied to content recommendation (such as agricultural knowledge recommendation, library book recommendation and the like) in other fields besides safe production knowledge recommendation, and the specific application method can be adjusted according to actual requirements, so that the invention is not limited.
When user data information is collected, the front-end embedded point mode is generally adopted to report clicking, searching, collecting and other behaviors of a user, historical behavior data of the user are obtained, and basic information (such as gender, age, region and the like) of the user is obtained. The page data is collected, for example by javascript, and sent to a user data collection server, which stores the data onto an HDFS (Hadoop Distributed File System, distributed file system).
It should be understood that the specific method for collecting user data information is only one specific example of the present invention, and other data collecting methods may be used in practical application of the present invention, which is not limited thereto.
After the user data information is obtained, the data is cleaned and data mined, and the interest model determining unit 610 is configured to mine characteristics such as preference, personal information and habit of the user based on the user data information, and construct a user interest model through a scoring matrix model.
For example: based on the user-interest scoring matrix representation, the user interest model is represented by an m x n matrix, where m is the number of system users and n is the user interest. User interest model R ij The interest score of user i for interest j is represented.
Fig. 2 is a flow chart of a personalized recommendation method for safety production knowledge, as shown in fig. 2, in the step of determining a user interest model (part a in the figure), user data information is obtained from a database or a cache, firstly, behavior features of a user are extracted according to user historical behavior data (a user behavior table), and feature vectors capable of representing the user interest are determined by combining with basic information of the user, and the user interest model is determined according to the feature vectors.
After determining the user interest model, a tag model determining unit 620 is configured to tag the user interest model according to a tag rule, and determine a user tag model. It can be understood that the specific type of the tag and the tag rule adopted for labeling the user interest model can be set according to actual requirements, and the invention is not limited to this.
The labels corresponding to the users can form the virtual models of the users, and further, the user groups can be subdivided through the labels, and different contents are recommended to the users of different virtual models when the users are individually pushed.
After determining the user tag model, the recommended content determining unit 630 is configured to search, based on the user tag model, for content matching the user tag model in the tagged security production knowledge base as the recommended content of the target security production knowledge.
It can be understood that each content file in the tagged security production knowledge base should include at least one corresponding tag, and the tag determination method in the tagged security production knowledge base may be determined according to actual requirements, which is not limited in the present invention.
According to the personalized recommendation system for the safety production knowledge, the user interest model is determined through the user historical behavior data and the user basic information, and the model is labeled to determine the user label model. And determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base according to the user label model. Through analysis and tagging of user interests, the interested safe production knowledge content is accurately recommended to the user in a personalized way, useless information browsing by the user is avoided, user experience is effectively improved, and user viscosity is enhanced.
It should be noted that, the personalized recommendation system for safety production knowledge provided by the present invention is used for executing the personalized recommendation method for safety production knowledge, and the specific embodiment and the method embodiment of the personalized recommendation system for safety production knowledge are consistent and are not described herein.
Fig. 7 is a schematic diagram of an entity structure of an electronic device according to the present invention, as shown in fig. 7, the electronic device may include: processor 710, communication interface 720, memory 730, and communication bus 740, wherein processor 710, communication interface 720, memory 730 communicate with each other via communication bus 740. Processor 710 may invoke logic instructions in memory 730 to perform a secure production knowledge personalized recommendation method comprising: determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information; labeling the user interest model and determining a user label model; and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
Further, the logic instructions in the memory 730 described above may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method of the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a Read-only memory (ROM), a random access memory (RAM, randomAccessMemory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method of personalized recommendation of secure production knowledge provided by the above methods, the method comprising: determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information; labeling the user interest model and determining a user label model; and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the above-provided security production knowledge personalized recommendation methods, the method comprising: determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information; labeling the user interest model and determining a user label model; and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
The apparatus embodiments described above are merely illustrative, wherein elements illustrated as separate elements may or may not be physically separate, and elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solutions may be embodied essentially or in part in the form of a software product, which may be stored in a computer-readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the various embodiments or methods of some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A personalized recommendation method for safe production knowledge, comprising:
determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information;
labeling the user interest model, and determining a user label model;
and determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
2. The personalized recommendation method for security production knowledge according to claim 1, further comprising, before the step of determining the recommended content of the target security production knowledge in a tagged security production knowledge base based on the user tag model:
determining a content tag based on the content information subject and the information keyword in the secure production knowledge base;
and based on the content label, mapping and binding the content information in the safe production knowledge base with the content label, and determining a labeled safe production knowledge base.
3. The personalized recommendation method for safety production knowledge according to claim 1, wherein the determining a user interest model based on user data information specifically comprises:
Analyzing the user data information according to a preset analysis direction to determine user attributes; wherein, the preset analysis direction includes: RFM analysis, user behavior analysis, user preference analysis, and user basic attribute analysis; the user attributes include: user natural attributes, user viscosity attributes, user preference attributes, user collection attributes, user search attributes, and activity policy attributes;
based on the user attributes, a user interest model is determined.
4. A method for personalized recommendation of security production knowledge according to any of claims 1-3, wherein said tagging said user interest model to determine a user tag model comprises:
labeling the user interest model according to a preset labeling rule, and determining a user label model; the user tag model comprises content information tags of interest to the user and association frequencies of the user and each tag.
5. The personalized recommendation method for security production knowledge according to claim 4, wherein determining the recommendation content of the target security production knowledge in the tagged security production knowledge base based on the user tag model, further comprises:
Determining a content list of interest of a user in the tagged security production knowledge base based on the user tag model;
and determining the recommended content of the target safe production knowledge according to a preset content recommendation rule based on the content list of interest of the user.
6. The personalized recommendation method for security production knowledge according to claim 5, wherein determining the recommendation content of the target security production knowledge in the tagged security production knowledge base based on the user tag model specifically comprises:
based on the attribute of the content information tag, determining the content of interest of the user in the tagged security production knowledge base according to a mixed recommendation algorithm; the mixed recommendation algorithm comprises a collaborative filtering recommendation algorithm and a content-based recommendation algorithm;
and ordering the content of interest of the user according to the association frequency of the user and each tag, and determining a content list of interest of the user.
7. The personalized recommendation method for safety production knowledge according to claim 5, wherein determining the recommended content of the target safety production knowledge based on the content list of interest of the user according to a preset content recommendation rule specifically comprises:
According to the user interested content list and a preset content recommendation rule, calculating the weight of the user interested content;
determining a recommendation candidate list according to the weight of the interesting content;
and determining the recommended content of the target safety production knowledge according to the recommended candidate list.
8. A personalized recommendation system for secure production knowledge, comprising: the system comprises a first model determining unit, a second model determining unit, a target word stock determining unit and a recommended content determining unit;
the first model determining unit is used for an interest model determining unit, a label model determining unit and a recommended content determining unit;
the interest model determining unit is used for determining a user interest model based on the user data information; wherein the user data information includes: user history behavior data and user base information;
the label model determining unit is used for labeling the user interest model and determining a user label model;
the recommended content determining unit is used for determining the recommended content of the target safe production knowledge in the labeled safe production knowledge base based on the user label model.
9. An electronic device comprising a memory and a processor, said processor and said memory completing communication with each other via a bus; the memory stores program instructions executable by the processor, the processor invoking the program instructions capable of performing the secure production knowledge personalized recommendation method of any of claims 1 to 7.
10. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor implements the security production knowledge personalized recommendation method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310131698.XA CN116431895A (en) | 2023-02-15 | 2023-02-15 | Personalized recommendation method and system for safety production knowledge |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310131698.XA CN116431895A (en) | 2023-02-15 | 2023-02-15 | Personalized recommendation method and system for safety production knowledge |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116431895A true CN116431895A (en) | 2023-07-14 |
Family
ID=87093257
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310131698.XA Pending CN116431895A (en) | 2023-02-15 | 2023-02-15 | Personalized recommendation method and system for safety production knowledge |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116431895A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117057959A (en) * | 2023-08-21 | 2023-11-14 | 武汉博晟安全技术股份有限公司 | Man-machine interaction type safety training method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886054A (en) * | 2014-03-13 | 2014-06-25 | 中国科学院自动化研究所 | Personalization recommendation system and method of network teaching resources |
| CN107122399A (en) * | 2017-03-16 | 2017-09-01 | 中国科学院自动化研究所 | Combined recommendation system based on Public Culture knowledge mapping platform |
| CN111966913A (en) * | 2020-10-21 | 2020-11-20 | 拼说说(深圳)网络科技有限公司 | Education resource recommendation processing method and device and computer equipment |
| WO2021249249A1 (en) * | 2020-06-11 | 2021-12-16 | 京东方科技集团股份有限公司 | Content recommendation method, apparatus and system |
| CN114493686A (en) * | 2022-01-14 | 2022-05-13 | 中国工商银行股份有限公司 | Operation content generation and pushing method and device |
-
2023
- 2023-02-15 CN CN202310131698.XA patent/CN116431895A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886054A (en) * | 2014-03-13 | 2014-06-25 | 中国科学院自动化研究所 | Personalization recommendation system and method of network teaching resources |
| CN107122399A (en) * | 2017-03-16 | 2017-09-01 | 中国科学院自动化研究所 | Combined recommendation system based on Public Culture knowledge mapping platform |
| WO2021249249A1 (en) * | 2020-06-11 | 2021-12-16 | 京东方科技集团股份有限公司 | Content recommendation method, apparatus and system |
| CN111966913A (en) * | 2020-10-21 | 2020-11-20 | 拼说说(深圳)网络科技有限公司 | Education resource recommendation processing method and device and computer equipment |
| CN114493686A (en) * | 2022-01-14 | 2022-05-13 | 中国工商银行股份有限公司 | Operation content generation and pushing method and device |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117057959A (en) * | 2023-08-21 | 2023-11-14 | 武汉博晟安全技术股份有限公司 | Man-machine interaction type safety training method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11663254B2 (en) | System and engine for seeded clustering of news events | |
| Kim et al. | A scientometric review of emerging trends and new developments in recommendation systems | |
| US10755179B2 (en) | Methods and apparatus for identifying concepts corresponding to input information | |
| US8832091B1 (en) | Graph-based semantic analysis of items | |
| CN109918563B (en) | Book recommendation method based on public data | |
| CN110175895B (en) | Article recommendation method and device | |
| Amami et al. | A graph based approach to scientific paper recommendation | |
| Kim et al. | A framework for tag-aware recommender systems | |
| Bouadjenek et al. | Persador: personalized social document representation for improving web search | |
| Velásquez | Web site keywords: A methodology for improving gradually the web site text content | |
| US20180089193A1 (en) | Category-based data analysis system for processing stored data-units and calculating their relevance to a subject domain with exemplary precision, and a computer-implemented method for identifying from a broad range of data sources, social entities that perform the function of Social Influencers | |
| Hutterer | Enhancing a job recommender with implicit user feedback | |
| Sun et al. | A hybrid approach for article recommendation in research social networks | |
| CN116431895A (en) | Personalized recommendation method and system for safety production knowledge | |
| Ferdaous et al. | Recommendation using a clustering algorithm based on a hybrid features selection method | |
| Rajkumar et al. | Users’ click and bookmark based personalization using modified agglomerative clustering for web search engine | |
| Lopes et al. | A personalized recommender system for digital libraries | |
| Servajean et al. | Profile diversity for query processing using user recommendations | |
| Yu et al. | Friend recommendation mechanism for social media based on content matching | |
| Khelghati | Deep web content monitoring | |
| Lucchese et al. | Recommender Systems. | |
| Wang et al. | Probabilistic attribute mapping for cold-start recommendation | |
| Katal et al. | Computational Intelligence Techniques for Recommendation System in Big Data | |
| Pillai | Query Recommendation System in Social Networks | |
| Rai et al. | Mining E-commerce Websites to Provide Efficient Methodology for Sentiment Analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |